The Effect of Motion Correction on Quantitative Perfusion Indices and Threshold-Based Interpretation in 13N-Ammonia Dynamic PET Myocardial Perfusion Imaging

Background/Objectives: The effect of cardiac motion on dynamic positron emission tomography (PET) myocardial perfusion imaging (MPI) could distort the quantitative perfusion indices during acquisition, depending on direction, magnitude, and vascular territory. This study aimed to assess the impact of motion correction (MC) and its influences on quantitative perfusion indices using a comprehensive range of established statistical methods. Method: We retrospectively analyzed 171 patients’ data who underwent 13N-ammonia dynamic PET/CT-MPI. All the rebinned listmode data were transferred to the dedicated workstation for MC and quantification, yielding paired MC and non-motion correction (NMC) datasets for both phases. Frame-by-frame post-MC vectors were used to characterize motion magnitude and direction. Paired comparisons, directional analysis, association, agreement, accuracy, precision, and clinical reclassification were performed. Results: Motion occurred during both acquisition phases, but it was more frequent, more heterogeneous, and generally greater during stress, with the most notable displacement along the Y-axis. MC had a limited impact on the rest myocardial blood flow (MBF) but showed more significant effects on stress MBF, myocardial flow reserve (MFR), and non-invasive fractional flow reserve (niFFR) -like index, especially in the left anterior descending artery (LAD). Although MC and NMC values remained strongly associated for MBF and MFR, they were not fully interchangeable. The bidirectional reclassification of MFR and niFFR-like index showed that MC responded to the direction and magnitude of motion rather than a simple increase the values. Conclusions: MC exerts a structured, bidirectional, and motion-dependent influence on quantitative dynamic PET-MPI and has an impact beyond simple numerical adjustment.